English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Fast gradient sign attack series are popular methods that are used togenerate adversarial examples. However, most of the approaches based on fastgradient sign attack series cannot balance the indistinguishability andtransferability due to the limitations of the basic sign structure. To addressthis problem, we propose a method, called Adam Iterative Fast Gradient TanhMethod (AI-FGTM), to generate indistinguishable adversarial examples with hightransferability. Besides, smaller kernels and dynamic step size are alsoapplied to generate adversarial examples for further increasing the attacksuccess rates. Extensive experiments on an ImageNet-compatible dataset showthat our method generates more indistinguishable adversarial examples andachieves higher attack success rates without extra running time and resource.Our best transfer-based attack NI-TI-DI-AITM can fool six classic defensemodels with an average success rate of 89.3% and three advanced defense modelswith an average success rate of 82.7%, which are higher than thestate-of-the-art gradient-based attacks. Additionally, our method can alsoreduce nearly 20% mean perturbation. We expect that our method will serve as anew baseline for generating adversarial examples with better transferabilityand indistinguishability.

Making Adversarial Examples More Transferable and Indistinguishable